DP Empirical Determinants of Bargaining Power RIETI Discussion Paper Series 06-E-030 UCHIDA Hirofumi

advertisement
DP
RIETI Discussion Paper Series 06-E-030
Empirical Determinants of Bargaining Power
UCHIDA Hirofumi
Wakayama University
The Research Institute of Economy, Trade and Industry
http://www.rieti.go.jp/en/
RIETI Discussion Paper Series 06-E -030
Empirical Determinants of Bargaining Power†
Hirofumi Uchida ‡
Faculty of Economics
Wakayama University
This version: May 1, 2006
Abstract
This paper empirically investigates what determine bargaining power between a
lender and a borrower who have continuing transactional relationships. Bargaining power
is proxied by which side of the transaction, i.e. the lender or the borrower, usually incurs a
shoe-leather cost when they have contact. The proxy is regressed on three types of
variables that can potentially determine distribution: (i) lender’s competition, (ii) the degree
of informational asymmetry between the two parties, and (iii) borrower performance.
Consistent with theoretical predictions, we find that intensive lender competition and
borrowers’ good performance increase the likelihood of the lender incurring the cost, or the
borrower’s power. We also obtain evidence suggesting that some lenders enjoy a status of
informational monopoly and capture borrowers.
Keywords: Bargaining power, Distribution, Banking relationship, Asymmetric information,
Competition
JEL classification code: D30, C78, L13, G21, L14
1. Introduction
The aim of this paper is to empirically investigate what determine bargaining power, or
distribution, between two parties involving a transaction.
From a uniquely designed survey of
corporate finance issues that contains information about continuing transactional relationships
between small- and medium-sized enterprises (SMEs) and their main lending banks, we obtain a
proxy for the bargaining power: which side of the transaction, a lender or a borrower, usually
†
This study was conducted as one of the projects of the Regional Finance Workshop of the Research
Institute of Economy, Trade, and Industry (RIETI). The author would like to thank RIETI for financial
support, Shuichi Uemura for administrative help and Nobuyoshi Yamori for designing and organizing the
survey. The author is also grateful to Shinsuke Kambe, Takashi Shimizu, Yuji Hosoya, Masaru
Yoshitomi, Yasuo Tanabe, and the seminar participants at the RIETI for useful comments. The author
respectfully dedicates this paper to his late supervisor Shoichi Royama, the discussion with whom the
original idea of this study is based on.
‡
Faculty of Economics, Wakayama University, 930 Sakaedani, Wakayama 640-8510, Japan. Tel.:
81-73-457-7710, Fax.: 81-73-457-7711, E-mail: uchida@eco.wakayama-u.ac.jp.
1
incurs a shoe-leather cost when they have contact. 1
We regress this unique proxy on three
types of variables that can potentially determine rent distribution: (i) lender’s competition (the
degree of competition), (ii) borrower performance (creditworthiness of the borrower), and (iii)
the availability of borrower’s information (the degree of informational asymmetry between the
two parties).
Consistent with theoretical predictions, we find that intensive lender competition and
borrowers’ good performance make the borrower powerful and increase the likelihood of lender
incurring the cost.
As for the availability of borrower’s information, although we find that
frequent availability of hard information decreases the likelihood, neither lender’s knowledge
about the borrower nor auditing of financial statements makes difference in determining the
bargaining power.
Thus, the impact of asymmetric information on bargaining power is found
to be relatively weak.
We also obtain evidence suggesting that some lenders enjoy a status of
informational monopoly and capture borrowers.
The biggest contribution of this paper is its uniqueness in directly testing theoretical
predictions from the theory of competition, the bargaining theory, or the economics of
information on the determinants of distribution.
confirm the validity of those theories.
On balance, the results obtained in this paper
Competition and borrower quality matter as
determinants of bargaining power, whereas informational asymmetry does not play a big role.
Distribution between a lender and a borrower is an important empirical issue in the
literature on banking relationships.
One of their focuses is on whether or not a stronger
bank-borrower relationship leads to any material benefit to the borrower in the form of
favorable terms of credit such as a low loan interest rate. 2
Although our focus here as
explained above is not limited to the effect from a strong banking relationship, since we have a
proxy for the strength of the banking relationship as an independent variable, the present study
can also be regarded as one of the benefit studies of banking relationship.
As such, the present
paper adds to this literature by making use of our unique proxy for the relationship benefit.
An
advantage of this proxy over the credit-term proxies in the conventional studies is that our proxy
captures an average benefit with a longer-term perspective in a sense that lender-borrower
1
In the case of bank lending to SMEs, lenders have frequent contact with the borrower in the course of a
long-term relationship in order to monitor and collect information about the borrowers’ creditworthiness.
For this strong banking relationship, see a survey by Boot (2000), for example.
2
See studies such as Petersen and Rajan (1994) and Berger and Udell (1995). See Boot (2000) as well.
2
contact is more frequent phenomena than loan contracting.
Burdett et al. (1995) theoretically investigate a closely related issue of should I stay or
should I go in a search model.
Their focus is however on an endogenous choice of go or stay
by agents searching for a new partner to transact.
In the present paper, we investigate the
distribution between two parties which already have some stable transactional relationships in
the form of the main bank relationship.
The remaining part of the paper is composed as follows.
In the next section, we briefly
survey theoretical predictions on the determinants of the bargaining power.
methodology are introduced in Section 3.
Section 4 reports the results.
Data and
The final section
concludes the paper.
2 Theory
In this paper, the term bargaining power refers in a broad sense to what determine the
“share of the cake,” or distribution, between two parties involving in a transaction.
In a Nash
bargaining context, the bargaining power is just defined as the relative weights put on utilities of
the trade partners which constitute their joint objective function to maximize.
The basic
Rubinstein alternating-offers model shows that the weights are determined by the relative
magnitude of the two players’ discount rates. 3
The term is, however, often used in different
contexts to refer to rent distribution in general between two parties involving a transaction.
We use the term in this broader sense.
As such, theories make several predictions about the determinants of the bargaining power.
First of all, we expect that the degree of competition is an important factor to determine the
power.
Theory of competition predicts that competition gets rid of excessive rents from the
competing parties.
Competition on one side of the transaction provides a party on the other
side with a greater chance to meet a different trade partner at a lower price.
We thus expect
that the more intensive competition that the lender faces, the greater power the borrower
obtains.
Borrower’s performance or creditworthiness is also a factor affecting the bargaining power.
If there is scarcity of supply in good borrowers, a creditworthy borrower attracts a bigger
3
If there is a risk of breakdown, the weights also depend on the breakdown probability. See Mothoo
(1999) for example.
3
number of lenders and thus obtains greater power.
The hypothesis is therefore that a more
creditworthy borrower obtains greater power.
Finally, informational asymmetry may also play a big role in the determination of the
bargaining power.
As the economics of information predicts in a variety of models, a party
with informational advantage can extract informational rent from a trading partner who has
inferior information.
This implies that as a lender becomes more knowledgeable about a
borrower, the lender obtains greater power.
However, we should also take into account the informational asymmetry among lenders
with respect to the information of the borrower’s creditworthiness.
If the knowledge the
incumbent bank obtains is private information, the bank obtains a status of informational
monopoly as opposed to competing banks, and can obtain greater power (Greenbaum et
al.(1989), Sharpe(1990), Rajan(1992), and von Thadden(2004)). However, if the knowledge is
public information which even competing banks can be accessible to, competition gets rid of the
power from the lender.
3 Data and methodology
Data source
The data source of this paper is the “Management Survey of Corporate Finance Issues in
the Kansai Area,” which was conducted in Japan in June 2005 by the Regional Finance
Workshop in the Research Institute of Economy, Trade, and Industry (RIETI).
This RIETI
survey asks small- and medium-sized enterprises (SMEs) about firm characteristics,
management strategy, bank relationships, and loan screening process. 4
We have effective
answers from 2020 firms which locate in the three prefectures, Osaka, Hyogo, and Kyoto, in the
Kansai area in Japan. 5 The average sample firm was established in 1967, employs 150 persons,
has a capitalization of 764 million yen, and has 1.19 billion yen of sales.
The length of the
relationship between an SME and its main bank in our data set is very long and is 27.9 years on
average.
4
The data set is the same as that for Uchida, Udell, and Yamori (2006a, 2006b). See Uchida, Udell, and
Yamori (2006a) for more details of the data set.
5
Kansai area is located in the middle of the main island of Japan, and the three prefectures in the area
form the focal point for the economy of western Japan.
4
Methodology and dependent variable
We run a regression with taking a proxy for bargaining power as a dependent variable.
The dependent and independent variables are explained below.
shown in Table 1.
Their descriptive statistics are
The availability of all the variables reduced the number of observations and
the regression is run with 1075 observations.
The dependent variable is a dummy variable THEYCOME, which is constructed as follows.
In the RIETI survey, firms are asked “where do you (= the firm’s entrepreneur or a person in
charge) usually meet with the loan officer of your main bank,” and the firm chooses an answer
from three options: 1. at your company, 2. at the bank branch office, and 3. other places.
The
variable THEYCOME takes a value of one if the firm chose option 1, and zero if the firm chose
option 2. 6
This variable represents which side of the transaction, a lender or a borrower,
usually incurs a shoe-leather cost when they have contact.
the lack of bargaining power.
Incurring the cost corresponds to
We run a Probit regression with this variable as the dependent
variable.
Main Independent variables
Based on the discussion in Section 2, we use independent variables that are grouped into
three categories.
First, to proxy bank competition, we use two variables.
The RIETI survey
asks borrowers, “whether the frequency of visit by banks other than the main bank increased in
the past two years.”
We construct a dummy variable VISITINC_OTHER which takes on a
value of one if the firm answered yes to this question. This dummy variable represents the
intensification of competition between the main bank and other banks (from a borrowers’ point
of view).
We also include the number of the banks that the firm currently borrows from,
N_BK, as an independent variable.
This variable captures the degree of competition among
banks which already have transactional relationships with the borrower.
We predict that these
variables have a positive impact on THEYCOME.
Second, we use two proxies for borrowers’ performance.
PERFORMANCE_S is a
dummy variable representing that the borrowing firm was in surplus (i.e. profitable) in the last
accounting year.
A dividend dummy DIVIDEND, which takes on a value of one if the firm
paid dividend in the last accounting year, also represents the firm’s good performance.
6
Sample firms which chose “other places” are excluded from the sample.
5
These
two variables proxy for good performance of the borrower, and should have a positive impact
on the dependent variable.
Finally, several variables are used as proxies for the availability of borrower’s information.
Here, we make a distinction between hard verifiable information from financial statements and
soft non-verifiable information such as the entrepreneur’s personal skills and integrity.
The
former can be transferred or documented, whereas the latter is difficult to document and
deteriorates through its transmission among agents or within organizational hierarchy.
This
distinction is important because they might have different impact on rent distribution.
The
former is likely to be public information, the availability of which should increase borrower
power.
The latter is likely to be private information, the production of which should increase
lender power (under intensive lender competition).
The distinction is important also because
different lending technologies might be employed depending on the hardness/softness of
information. 7
To represent availability of hard information, we use three variables.
on a value of one if the firm has its financial statements audited by a CPA.
First, AUDIT takes
Second, FINSTAT
is a dummy variable which equals to one if the firm has any financial statements.
Finally,
FREQ_FINSTAT represents how many times a year the firm submits financial statements to the
main bank.
Among these three variables, AUDIT is the most important because it represents
that the information from financial statements is certified by a third party, and is thus reliable.
If these variables capture publicly available information, they are likely to have a positive
impact on THEYCOME.
However, if hard information proxied by these variables is private
information, the impact should be negative.
The availability of soft information is difficult to measure.
We thus use an index a la
Scott (2004) that captures to what extent the main bank is knowledgeable about the borrower in
terms of soft information.
The RIETI survey asks the respondent firms about characteristics
regarding which the firm rates the lender’s knowledge.
information. 8
Six of them are with respect to soft
Following Scott (2004) and using these six characteristics, we created a measure
7
When hard information from financial statements is available, banks can deploy the financial statement
lending technology. When only soft information is available, banks should employ the relationship
lending technology. See Berger and Udell (2002, 2006) for more on lending technologies.
8
The six characteristics are: 1. [The lender] Knows your business (the firm itself and its business), 2.
[The lender] Knows your managers and owners (the firm's managers and owners), 3. [The lender] Knows
6
of soft information production, SOFT. 9
Since SOFT is likely to be associated with private
knowledge of the main bank, we expect that the variable has a negative impact on
THEYCOME.
Relationship strength
LENGTH is the length of the relationship between the borrower and the main bank.
Since the main bank can accumulate proprietary information about the borrower through a
long-term relationship, this variable may have a negative impact on THEYCOME.
However,
if the information is not unique to the main bank or is gradually funneled out, the bank cannot
obtain a monopoly status and then LENGTH may have a positive impact.
It should be noted here that because THEYCOME proxies for a relative benefit between a
lender and a borrower, the regression with LENGTH as an independent variable can be
considered as an analysis on whether there is any benefit from a strong bank-borrower
relationship.
There is abundance of studies on the banking relationships which empirically
examine the benefit by using terms of credit such as a loan interest rate as proxies for the
benefit. 10
As one of such studies, the present study adds to the literature by using our unique
proxy for the benefit.
Compared with frequent lender-borrower contacts in the course of a
long-term relationship, loan contracts are made occasionally, so that credit terms are observed
relatively infrequently.
Our unique proxy, i.e. which usually incurs a cost of contact, has an
advantage over conventional credit-term proxies since it captures benefit on a continuous basis
in the course of everyday contact between the lender and the borrower.
Individual credit terms
may not be good proxies also because they are simultaneously determined and are thus difficult
to evaluate in their isolation.
your industry (the firm's industry), 4. [The lender] Knows the local community (the local community the
firm is in), 5. [The lender] Know the market (the firm's market), and 6. Social contact with loan officer
(loan officer has frequent contact with the firm).
9
With respect to each of the characteristics, the survey asks (1) how important the firm reckons it in
doing business with a financial institution (in general). This question captures the firm’s perception of
what define a good lender. The firms are also asked to evaluate (2) to what extent the current main bank
is satisfactory for each characteristic. This captures the firm’s evaluation of the current main bank.
The answers to the both questions are chosen on a 5-point scale from 1 (very good) to 5 (very bad). For
each of the six characteristics, we constructed a dummy variable which takes a value of one if the firm
chose 1 for both questions (1) and (2). SOFT is the first component of the principal component analysis
over the resulting six dummies. Note that the survey that Scott (2004) uses allows him to use only four
similar characteristics to construct an index.
10
See Petersen and Rajan (1994), Berger and Udell (1995), and Kano et al. (2006), for example.
7
Control variables
We also use different control variables.
ASSET is the size of the firm’s asset, FIRMAGE
is the age of the firm, EMPLOYEE is the number of employees.
These variables may also
represent the size-related availability of hard and public information of the borrower that is
different from financial statement information.
should be negative.
In that case, the coefficients of these variables
Similar prediction applies to other two dummy variables, LISTED which
represents that the firm is listed, and AFFILIATE which represents that the firm is affiliated
with a parent company.
We also use two bank-type dummies to represent the size and the organizational
complexity of the lending bank.
COOPERATIVE takes on a value of one if the main bank is
either a Shinkin bank or a credit cooperative.
REGIONAL represents that the firm’s main
bank is either a regional bank or a second tier regional bank.
Banks represented by
COOPERATIVE is smaller in size and less complex in terms of organizational structure than
banks represented by REGIONAL.
The latter is however smaller and less complex than a
bank of three types (= default): a city bank, a long-term credit bank, or a trust bank.
These
dummies are introduced to control for the difference in lenders’ characteristics and the lending
technologies used.
Stein (2002) theoretically shows that while small banks are good at
producing soft information, large banks are not.
This implies a negative coefficient of
COOPERATIVE (and REGIONAL to a lesser extent).
Finally, we use industry dummies and urban (regional) dummies.
Industry dummies are
CONST (construction), IT (information and telecommunication), TRANSPORT (transportation),
WHOLESALE (wholesale), RETAIL (retail), REALESTATE (real estate), and SERVICES
(other services).
Three urban dummies are OSAKA, KOBE, and KYOTO which are
constructed from the firm’s area code of the telephone number and thus represent that the firm is
located in or around the three biggest cities in the Kansai area.
The default firm is in the
manufacturing industry and locates in rural areas.
4. Results
Table 2 reports the regression results of the Probit estimation.
First, VISITINC_OTHER
which represents increased competition among lenders is positive and significant, meaning that
8
intensification of lenders’ competition leads to lenders incurring the shoe-leather cost.
consistent with a theoretical prediction that competition gets rid of rents.
lending banks is however not significant.
This is
The number of
This may imply that it is not competition with
incumbent banks but competition with potential lenders that matters.
As for the two performance variables, both of them have a positively significant impact on
THEYCOME.
Consistent with a theoretical prediction, lenders tend to incur the shoe-leather
cost and visit borrowers when their performance is good.
Creditworthy borrowers have bigger
bargaining power.
Among the four information availability variables, only FREQ_FINSTAT is significant.
Frequent submission of financial statements contributes to the borrowers’ visits to the lenders’
place.
This is consistent with the theoretical prediction in a sense that increased transparency
of borrowers reduces lenders’ informational disadvantage and thereby increases lender power.
However, this result may also imply that the borrower with small bargaining power has to
disclose its information more frequently.
On balance, the results on information availability
variables are not very strong since the other three variables are insignificant.
Finally, LENGTH has a negative impact on THEYCOME.
This may imply that a bank
can accumulate information about the borrower through a longer relationship and obtain a status
of informational monopoly.
The borrower is then captured by the bank and has to incur a
shoe-leather cost.
5. Conclusion
Using a unique proxy, we examined the determinants of bargaining power between lenders
and borrowers.
The results were consistent with theoretical predictions that competition
among lending banks and good performance of the borrower contribute to the borrowers’
extraction of more bargaining power from the lender.
We also found evidence that resolving
informational asymmetry leads to lenders obtaining more bargaining power, although the
evidence was weak.
On balance, the results obtained in this paper confirm the validity of those theories.
Competition and borrower quality matter as determinants of bargaining power, although
informational asymmetry does not play a very big role.
9
References
A.N. Berger and G. Udell. Relationship lending and lines of credit in small firm finance. Journal of
Business, Vol. 68, pp. 351–381, 1995.
A.N. Berger and G.F. Udell. Small business credit availability and relationship lending: The importance of
bank organizational structure. Economic Journal, Vol. 112, pp. F32–F53, 2002.
A.N. Berger and G.F. Udell. A more complete conceptual framework for SME finance. Journal of
Banking and Finance, forthcoming 2006. .
A.W.A. Boot. Relationship banking: what do we know? Journal of Financial Intermediation, Vol. 9, pp.
7–25, 2000.
K. Burdett, M. Coles, N. Kiyotaki, and R. Wright. Buyers and sellers: Should I stay or should I go?
American Economic Review, Papers and Proceedings, vol.85, pp. 281-286, 1995.
S.I. Greenbaum, G. Kanatas, and I. Venezia. Equilibrium loan pricing under the bank-client relationship.
Journal of Banking and Finance, Vol. 13, pp. 221–235, 1989.
M. Kano, H. Uchida, G.F. Udell, and W. Watanabe. Information Verifiability, Bank Organization, Bank
Competition and Bank-Borrower Relationships. RIETI discussion paper 06-E-003, the Research
Institute of Economy, Trade, and Industry, 2006.
A. Mothoo. Bargaining Theory with Applications. Cambridge University Press, 1999.
M.A. Petersen and R.G. Rajan. The benefits of lending relationships: evidence from small business data.
Journal of Finance, Vol. 49, pp. 3–37, 1994.
R. Rajan. Insiders and outsiders: The choice between informed and arm’s length debt. Journal of Finance,
Vol. 47, pp. 1367–1400, 1992.
J.A. Scott. Small business and the value of community financial institutions. Journal of Financial Services
Research, Vol. 25, pp. 207--230, 2004.
S.A. Sharpe. Asymmetric information, bank lending, and implicit contracts: A stylized model of customer
relationships. Journal of Finance, Vol. XLV, pp. 1069–1087, 1990.
10
The Small and Medium Enterprise Agency of the Government of Japan. 2002 White Paper on Small and
Medium Enterprises in Japan. 2002.
The Small and Medium Enterprise Agency of the Government of Japan. 2003 White Paper on Small and
Medium Enterprises in Japan. 2003.
The Small and Medium Enterprise Agency of the Government of Japan. 2004 White Paper on Small and
Medium Enterprises in Japan. 2004.
J.C. Stein. Information production and capital allocation: decentralized versus hierarchical firms, Journal
of Finance LVII, 1891–1921, 2002.
H. Uchida, G. Udell and N. Yamori. SME financing and the choice of lending technology, RIETI
discussion paper 06-E-025, Research Institute of Economy, Trade, and Industry, 2006a.
H. Uchida, G. Udell and N. Yamori. Loan officers and relationship lending, forthcoming as a RIETI
discussion paper, 2006b.
E.L. von Thadden. Asymmetric information, bank lending and implicit contracts: the winner’s curse.
Finance Research Letters, Vol. 1, pp. 11–23, 2004.
11
Table 1 Descriptive Statistics
Variable
THEYCOME
VISITINC_OTHER
N_BK
PERFORMANCE_S
DIVIDEND
FREQ_FINSTAT
FINSTAT
AUDIT
SOFT
LENGTH
LOG(ASSET)
LOG(EMPLOYEE)
FIRMAGE
LISTED
AFFILIATE
REGIONAL
COOPERATIVE
CONST
IT
TRANSPORT
WHOLESALE
RETAIL
REALESTATE
RESTAU_HOTEL
SERVICES
OSAKA
KOBE
KYOTO
Obs
1539
1649
1563
1649
1649
1444
1649
1649
1649
1514
1486
1637
1479
1649
1649
1649
1649
1649
1649
1649
1649
1649
1649
1649
1649
1649
1649
1649
Mean Std. Dev
Min
0.803
0.398
0
0.559
0.497
0
3.638
3.145
0
0.813
0.390
0
0.429
0.495
0
2.499
2.986
1
0.953
0.211
0
0.129
0.335
0
0.072
1.725 -0.7323
27.053
16.324
0
21.014
1.737 12.9480
3.951
1.280
0.0000
49.360
24.368
1
0.040
0.196
0
0.131
0.337
0
0.192
0.394
0
0.159
0.366
0
0.124
0.330
0
0.029
0.166
0
0.067
0.251
0
0.198
0.399
0
0.058
0.234
0
0.023
0.150
0
0.018
0.134
0
0.118
0.322
0
0.423
0.494
0
0.096
0.295
0
0.145
0.352
0
Max
1
1
34
1
1
13
1
1
9.1724
109
28.2955
10.0470
136
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Table 2 Estimation Result
(Independent variable = THEYCOME)
Variable
VISITINC_OTHER
Lender Competition
N_BK
PERFORMANCE_S
Borrower Performance
DIVIDEND
Soft Info. SOFT
Availability of
FREQ_FINSTAT
Information Hard Info. FINSTAT
AUDIT
LENGTH
LOG(ASSET)
LOG(EMPLOYEE)
Borrower Characteristics
FIRMAGE
LISTED
AFFILIATE
REGIONAL
Lender Type
UNION
CONST
IT
TRANSPORT
WHOLESALE
Borrower Industry
RETAIL
REALESTATE
RESTAU_HOTEL
SERVICES
OSAKA
Borrower Region
KOBE
KYOTO
Prob > chi2
Pseudo R2
Number of observations
dF/dx
0.0637
-0.0023
0.0646
0.0929
0.0067
-0.0214
-0.0705
-0.0002
-0.0023
0.0291
-0.0143
0.0006
0.0681
-0.1305
-0.0353
-0.0272
-0.0080
0.0594
0.0317
0.0725
0.0014
-0.1020
0.1093
0.0667
0.0409
0.0085
0.0144
0.0000
0.1194
1075
**
*
***
***
***
**
***
**
*
Std. Error
0.0252
0.0046
0.0339
0.0253
0.0068
0.0034
0.0511
0.0355
0.0008
0.0142
0.0182
0.0006
0.0609
0.0496
0.0335
0.0366
0.0370
0.0563
0.0435
0.0279
0.0488
0.1120
0.0513
0.0311
0.0268
0.0413
0.0339
Prob.
0.0100
0.6170
0.0410
0.0000
0.3240
0.0000
0.2640
0.9950
0.0070
0.0410
0.4330
0.3310
0.3580
0.0030
0.2730
0.4410
0.8270
0.3710
0.4960
0.0210
0.9770
0.2990
0.1870
0.0650
0.1320
0.8390
0.6770
Notes:
The result from a Probit estimation is shown. The dependent variable THEYCOME is
equal to one if the borrower and the main bank usually have contact at the borrower's
place as opposed to the branch office of the main bank.
The main independent variables are grouped into three categories: (1) Lenders
comeptition, (2) Availability of Information, and (3) Borrower Performance. (1) Lenders
comeptition: Dummy variables VISITINC_OTHER and VISITINC_MB represent the
increase in visits from the banks other than the main bank (OTHER) and the main bank
(MB), respectivelly. N_BK represents the number of lending banks. (2) Availability of
Information: SOFT is an index a la Scott (2002) for the lender's knolwedge about the
borrower in terms of soft information. FREQ_FINSTAT is the frequency that the
borrower submits its financial statements to the borrower. FINSTAT represents the
existence of any financial statements. AUDIT represents that the borrower have its
financial statements audited. (3) Borrower Performance: PERFORMANCE_S equals to
***, **, or * means that the coefficient is statistically significant at 1%, 5%, or 10%
level.
Download